Parcels-code/Parcels

Now that we have Parquet output, its really easy to integrate this into Kepler.gl for exploratory analysis. Its nice to use Matplotlib for visualisations in the notebooks, but I think it would be better to recommend users use Kepler for their EDA as its much more interactive, and scales up nicer. **Using the same data from https://docs.parcels-code.org/en/main/user_guide/getting_started/tutorial_output.html ** Kepler version: kepler.mp4 Matplotlib version: Screen.Recording.2026-08-31.at.14.42.51.mov

What version of Parcels are you running? main Is your feature request related to a problem? Currently notebooks and doc pages show cell inputs and outputs in their entirety. There are cases where we don't want this: Plotting code - we would want to use a hide-input cell tag to put the cell inputs in a collapsable block Code for cleaning up .parquet temporary files - we would want to use remove-input and remove-output cell tags to completely hide these cells This can be controlled via myst-nb cell tags, and gives us more control over what we draw attention to in our documentation. Describe the solution you'd like Applying of cell tags appropriately. Note generally that when working with myst formatting the web browser version of jupyter lab (i.e., via the jupyter lab local server) is better as it shows rich previews, compared to the vscode version of Jupyter which doesn't support myst. Describe alternatives you've considered No response Additional context No response

Hmm, this would also return True for a B-grid (where U and V are on the same node, but W and tracers are on different nodes). So the naming here is not correct. Perhaps change to _is_cgrid and then return the negate? Originally posted by @erikvansebille in #2646 (comment)

What version of Parcels are you running? v4 Is your feature request related to a problem? Running pytest tests/ now throws a lot of warnings (see list below for 481decc). Although a few of these are upstream (so we can't solve them), most of these are in our own code. It would be good to fix these Once the list is empty, we can implement #2413 to fail on warnings src/parcels/_datasets/remote.py:195 tests/test_advection.py::test_nemo_curvilinear_fieldset tests/test_convert.py::test_convert_nemo_offsets tests/test_index_search.py::test_indexing_nemo_curvilinear /Users/erik/Codes/parcels/src/parcels/_datasets/remote.py:195: FutureWarning: dropping variables using `drop` is deprecated; use drop_vars. return ds.isel(time=0).drop(["time"]) tests/sgrid/test_accessor.py: 214 warnings tests/test_advection.py: 12 warnings /Users/erik/Codes/parcels/src/parcels/_sgrid/accessor.py:101: FutureWarning: The return type of `Dataset.dims` will be changed to return a set of dimension names in f

What version of Parcels are you running? N/A Is your feature request related to a problem? Currently the ParticleSet.execute(...,verbose_progress=True) progress bar prints a new line to job logs if run from a non-interactive shell, e.g. from slurm. Describe the solution you'd like This could possibly be prevented by catching the type of shell and handling the progress bar accordingly. Describe alternatives you've considered No response Additional context No response